Elsevier

Clinical Nutrition

Volume 35, Issue 1, February 2016, Pages 158-162
Clinical Nutrition

Original article
Identifying critically-ill patients who will benefit most from nutritional therapy: Further validation of the “modified NUTRIC” nutritional risk assessment tool

https://doi.org/10.1016/j.clnu.2015.01.015Get rights and content

Summary

Introduction

Better tools are needed to assist in the identification of critically ill patients most likely to benefit from artificial nutrition therapy. Recently, the Nutrition Risk in Critically ill (NUTRIC) score has been developed for such purpose. The objective of this study was to externally validate a modified version of the NUTRIC score in a second database.

Methods

We conducted a post hoc analysis of a database of a randomized control trial of intensive care unit (ICU) patients with multi-organ failure. Data for all variables of the NUTRIC score with the exception of IL-6 levels were collected. These included age, APACHE II score, SOFA score, number of co-morbidities, days from hospital admission to ICU admission. The NUTRIC score was calculated using the exact same thresholds and point system as developed previously except the IL-6 item was omitted. A logistic model including the NUTRIC score, the nutritional adequacy and their interaction was estimated to assess if the NUTRIC score modified the association between nutritional adequacy and 28-day mortality. We also examined the association of elevated NUTRIC scores and 6-month month mortality and the interaction between NUTRIC score and nutritional adequacy.

Results

A total of 1199 patients were analyzed. The mean total calories prescribed was 1817 cal (SD 312) with total mean protein prescribed of 98.3 g (SD 23.6). The number of patients who received PN was 9.5%. The overall 28-day mortality rate in this validation sample was 29% and the mean NUTRIC score was 5.5 (SD 1.6). Based on the logistic model, the odds of mortality at 28 days was multiplied by 1.4 (95% CI, 1.3–1.5) for every point increase on the NUTRIC score. The mean (SD) nutritional adequacy was 50.2 (29.5) with an interquartile range from 24.8 to 74.1. The test for interaction confirmed that the association between nutritional adequacy and 28-day mortality is significantly modified by the NUTRIC score (test for interaction p = 0.029). In particular, there is a strong positive association between nutritional adequacy and 28 day survival in patients with a high NUTRIC score but this association diminishes with decreasing NUTRIC score. Higher NUTRIC scores are also significantly associated with higher 6-month mortality (p < 0.0001) and again the positive association between nutritional adequacy and 6 month survival was significantly stronger (and perhaps only present) in patients with higher NUTRIC score (test for interaction p = 0.038).

Conclusion

The NUTRIC scoring system is externally validated and may be useful in identifying critically ill patients most likely to benefit from optimal amounts of macronutrients when considering mortality as an outcome.

Introduction

Heyland et al. previously proposed a novel scoring tool, the Nutrition Risk in Critically ill (NUTRIC) score, which is the first nutritional risk assessment tool developed and validated specifically for intensive care unit (ICU) patients [1]. Many other risk scores and assessment tools exist to quantify nutrition risk [2], [3], [4], [5], [6], [7] but none have been specifically designed for ICU patients [7]. Indeed, they generally consider all critically ill patients to be at high nutritional risk [2], [8]. However, the recognition that not all ICU patients will respond the same to nutritional interventions was the critical concept behind the NUTRIC score [1], [8], [9]. The conceptual model incorporated candidate predictor markers of acute starvation, chronic starvation, acute inflammation and chronic inflammation [1], [9]. All candidate predictors incorporated into our final model predictors were significantly associated with 28-day mortality [1]. Measure of under-nutrition, such as history or reduced oral intake or recent weight loss, did not factor into the final model because of significant amounts of missing data. The final composite score accurately identified those patients who had higher mortality rates or survivors with longer lengths of stay. In addition, there was an interaction between mortality, nutritional intake and NUTRIC score suggesting that those with higher NUTRIC scores (6 or more) benefited the most from increasing nutritional intake. However, the inferences about the validity of the NUTRIC score are limited because they are derived and validated in the same database.

Many methods of nutritional screening in hospitalized patients are cumbersome and time-consuming and hence are not routinely done [10]. The NUTRIC score is easy to calculate as it contains variables that are mostly easy to obtain in the critical care setting, with the exception of IL-6 levels which is not commonly measured. In practice, many units are using the NUTRIC score without the IL-6 level and the question remains as to the validity of the validity of the NUTRIC score without IL-6 level (modified NUTRIC score). The second stage in development of a clinical ICU prediction model is external validation [11]. The aim of this study is to externally validate [11] this modified NUTRIC score in a second, population of critically ill patients. We hypothesize that the modified NUTRIC score will retain its validity in this new database by omitting the IL-6 levels, and we can increase the clinical utility of the tool.

Section snippets

Methods

This study was a post hoc analysis of an existing database derived from a randomized control trial conducted in 40 tertiary ICU's in Europe and North America, after ethics approval was obtained. The purpose of the trial was to evaluate the effectiveness of glutamine and antioxidant supplementation in critically ill patients [12]. All patients were attempted to be fed according to the Canadian Critical Care Nutrition practice guidelines, independent of study supplements [12]. The trial

Results

Five patients withdrew consent prior to treatment and were not evaluable for 28-day mortality, and the amount of calories received was not known for an additional 19 patients. Thus, the current analysis included 1199 patients.

The overall 28-day mortality rate in this validation sample is 29.0% compared to 23.1% in the NUTRIC development sample. The distribution of the items included in the NUTRIC risk score are presented for both this validation sample and the previous development sample in

Discussion

We set out to provide a second validation of the NUTRIC score in a second database and this time, without IL-6 levels. We report that a logistic model with NUTRIC score, excluding IL-6, as the sole continuous independent variable predicted mortality with odds of mortality multiplied by 1.4 (95% CI, 1.3–1.5) for every point increase on the NUTRIC score. We demonstrate that increased nutritional adequacy is associated with increased survival in patients with higher NUTRIC scores (≥6) but not in

Conclusion

We have demonstrated independent validation of the NUTRIC score without IL-6 levels to help discriminate which ICU patients will benefit more (or less) from early adapted protein-energy provision. This scoring tool represents the first nutritional risk assessment tool developed and validated specifically for ICU patients. The NUTRIC score is a practical, easy-to-use tool based on variables that are easy to obtain in the critical care setting. We assert that not all ICU patients are the same,

Conflict of interest

None declared.

References (17)

There are more references available in the full text version of this article.

Cited by (0)

View full text